This study will investigate the acute relationship between alcohol consumption and balance in elderly women. This is one component of a much broader and poorly understood area, the health risks associated with social drinking among elderly persons. This study, building on existing knowledge of associations between advanced age and balance impairment, and between advanced age and altered responses to alcohol, will seek to answer the questions of whether or not alcohol acutely impairs balance in elderly women who are social drinkers, and whether or not elderly women experience a greater degree of balance impairment than do young adult women after ingestion of alcohol. A double-blind, latin-square cross-over design with balanced treatment sequence orders is proposed. Subjects will be recruited by advertisement, and eligibility will be determined by applying strict drinking behavior and health criteria. Twelve white female social drinkers 70-79 years of age and twelve white female social drinkers 21-30 years of age will be enrolled, and randomization will be performed in each age group to equal- sized groups of treatment sequence orders. The four treatment conditions will be placebo, low dose alcohol, moderate dose alcohol, and high dose alcohol. Each subject will be tested four times, with intervals of approximately 2-3 days between testing sessions. On each testing day, subjects will be given a standard meal and will then be fasted for 2 and 1/2 hours. Each subject will then undergo balance testing and electronystagmographic testing before ingesting the beverage, and at set alcohol levels (or at set time intervals after placebo) after consuming the beverage. Balance testing will be performed with dynamic posturography, which is a computer-controlled moveable platform, with visual surround and safety harness, that precisely and accurately measures both the sensory and motor components of balance. Blood alcohol concentrations will be measured with breath analysis at 5 minute intervals. Scales with questions measuring expectancy of alcohol effects and perceived level of intoxication will also be administered. Repeated measures analysis of variance will be used to compare within and between groups, balance before, and at the specific blood alcohol concentrations after the beverages are consumed, to determine the relationship between blood alcohol concentrations and balance, and to determine the effects of age on that relationship. Multiple linear regression will be used to examine the relationship between alcohol and balance, and test for a dose- response relationship.